The study examined the interview performance of 310 applicants to the Metropolitan Police and then followed through training school and probation the 109 who subsequently joined the force. Their predicted performance at interview was compared with their actual performance at training school and on probation. The study was conducted from February 1983, when the interviews started, until October 1985, when the last joiners reached the end of their probation. The supervisors of the 109 candidates who eventually joined the force were asked to rate them on the same rating form used in the selection interview both at training school and during their probation at 8, 12, 18, 21, and 23 months of service. If the interviewers were accurate in their ratings, then their scores would correspond with the scores given by the supervisors. There was no significant correlation between the ratings of the recruits at interview and their ratings at training school or on probation. Interviewers are apparently being asked to perform a task for which they have not been properly equipped. Interviewers must be given some sort of framework for the interview so as to assess what skills and abilities are required to do a police officer's job. These requirements must be measurable. 3 tables, 8 references.